Preface

Reflecting on Prediction Fiascos: A New Year's Thought

As we step into a new year, it's a perfect time to reflect on the past and learn from it. One area worth pondering is the notable instances when pollsters and economic analysts have gotten their predictions terribly wrong. These moments of miscalculation can teach us valuable lessons about the nature of forecasting and the importance of critical thinking.

In the past year 2024, as in 2023, I was very intensively involved in international project missions - be it in Europe, Brazil, Argentina or Mexico/USMCA economic area. In many conversations, I was able to perceive trends that differed from those that the herds of business analysts and pollsters were broadcasting to the world.

Why Reflect on Prediction Failures?

Understanding Uncertainty. Predictions, whether in politics or economics, are inherently uncertain. Reflecting on past errors reminds us that no forecast is foolproof. This awareness can help us approach future predictions with a healthy dose of skepticism and caution.

Improving Methods. Analyzing why predictions failed can lead to better methodologies. For instance, after several high-profile polling errors, experts have revisited their models and data collection techniques to improve accuracy.

Managing Expectations. When we understand that predictions can be wrong, we can manage our expectations better. This is crucial for both individuals and businesses, helping them to prepare for a range of possible outcomes rather than relying on a single forecast.

Encouraging Critical Thinking. Reflecting on prediction fiascos encourages us to think critically about the information we consume. It reminds us to question sources, consider alternative viewpoints, and not take predictions at face value.

As we embrace the new year, let's take a moment to learn from the past. By understanding the limitations and potential pitfalls of predictions, we can navigate the future with greater wisdom and resilience.

In the world of economic forecasting, accuracy is paramount. Yet, in Brazil, both pollsters and specially business analysts have often missed the mark, leading to widespread frustration and skepticism. This phenomenon isn't just a minor hiccup; it's a glaring issue that has significant implications for businesses, investors, and the general public.

The Overly Optimistic Years

Let's rewind to a few years ago when Brazil's economic outlook was painted with broad strokes of optimism. Analysts predicted robust growth rates, fueled by a booming commodities market and a seemingly stable political environment. These forecasts, however, turned out to be overly enthusiastic. The reality was starkly different. Economic growth lagged far behind predictions, leaving businesses and investors grappling with unexpected challenges. Who remembers that at the end of 2018 and until at least mid-2019, Brazilian banks and analysts at Faria Lima in São Paulo were shouting loudly about a massive appreciation of the Brazilian real and an exchange rate of less than 3.0 to the US dollar?

What was the reality?

The massive devaluation trend began on the first trading day of the fourth quarter of 2019. Faria Lima wanted to forget that very quickly. The end of the year was approaching - people had other things on their minds. When the pandemic started early in 2020, the story could be rewritten.

It was the pandemic! Do not look back - it was the pandemic!

Too bad, anyone who is really interested in things will quickly see that the pandemic didn't even exist in 2019.

But never mind - analysts aren't interested in yesterday's gossip.

The Pendulum Swings: From Optimism to Pessimism

Fast forward to recent years, and the pendulum has swung to the opposite extreme. Now, analysts are predicting doom and gloom, forecasting dismal economic performance. Yet again, these predictions have proven to be off the mark. Brazil's economy, while not booming, has shown resilience and growth that defies the bleak forecasts. This flip-flopping between extremes has eroded trust in economic predictions and left many questioning the credibility of these so-called experts. It is enough to go back exactly one year. The analysts at Faria Lima and the Brazilian banks never tired of predicting maximum GDP growth of 1 or 1.2% in the best-case scenario. But growth would probably tend to be closer to zero in 2024, with the realized result being a plus of 3.6%.

How can you be so massively wrong?

The Consequences of Inaccurate Forecasting

The impact of these inaccurate forecasts is far-reaching. Businesses rely on economic predictions to make informed decisions about investments, hiring, and expansion. When forecasts are wildly off, it can lead to poor decision-making, financial losses, and missed opportunities. Investors, too, suffer as they navigate a market clouded by unreliable data. For the general public, these inaccuracies can lead to misplaced optimism or undue pessimism, affecting consumer confidence and spending.

Why Do They Get It So Wrong?

Several factors contribute to the frequent inaccuracies in economic forecasting in Brazil:

Complex Economic Environment. Brazil's economy is influenced by a myriad of factors, including political instability, global market fluctuations, and internal economic policies. This complexity makes accurate forecasting challenging.

Overreliance on Models. Many analysts rely heavily on economic models that may not fully capture the nuances of Brazil's unique economic landscape. These models often fail to account for unexpected variables and shocks. In virtually all cases, these are also extremely complex linear models, although business analysts should have realized by now that economic environments and economies are also non-linear realities.

Bias, Groupthink and Gossip. There is often a herd mentality among analysts, where initial optimistic or pessimistic forecasts are echoed by others, creating a feedback loop that amplifies inaccuracies. This is amplified by superficial ideological bias based on catchy keywords - nothing more. The WhatsApp screens are up and burning!

Moving Forward: A Call for Better Forecasting

It's clear that the current state of economic forecasting in Brazil needs a serious overhaul. Analysts must adopt more nuanced approaches, incorporating a wider range of data, create scenarios based on none-linear systems and considering the unique aspects of Brazil's economy. Transparency in methodology and a willingness to adjust predictions in light of new data are crucial steps toward rebuilding trust.

Let's Look Up - to the North!

The Polling Predicament: Misleading Predictions in the USA

In the realm of political forecasting, pollsters hold a significant influence over public perception and media narratives. However, recent election cycles in the United States have highlighted a troubling trend: pollster predictions often miss the mark, leading to widespread confusion and mistrust.

The 2020 Election: A Case Study in Error

The 2020 presidential election serves as a prime example of polling inaccuracies. Many polls predicted a decisive victory for Joe Biden, with some even suggesting a landslide. However, the actual results were much closer, with Donald Trump performing significantly better than expected. This discrepancy was not just a minor error; it was the largest polling error in 40 years. The overestimation of Biden's support by nearly 4 percentage points left many questioning the reliability of pre-election polls.

The 2024 Election: Improvement or More of the Same?

Fast forward to the 2024 election, and the story is both similar and different. While some improvements were noted, such as a closer alignment between polling averages and actual results in key swing states, significant errors persisted. Polls underestimated Trump's support once again, leading to another unexpected victory for him. This recurring underestimation of Trump's appeal suggests a fundamental flaw in how pollsters gauge voter sentiment.

Why Do Polls Get It Wrong?

Several factors contribute to the inaccuracies in polling:

Sampling Issues. Pollsters often struggle to obtain a representative sample of the electorate. Certain demographics, particularly those who are less likely to respond to surveys, can be underrepresented.

Social Desirability Bias. Voters may not always disclose their true preferences, especially if they feel their choice is socially unpopular. This can skew results, particularly in politically charged environments.

Late-Deciding Voters. A significant number of voters make their decision in the final days leading up to the election, which can be difficult for polls to capture accurately.

Weighting and Adjustments. Pollsters use various methods to adjust their data to reflect the electorate. However, these adjustments can sometimes introduce errors if the underlying assumptions are incorrect.

The Consequences of Misleading Polls

The impact of inaccurate polling extends beyond mere numbers. Misleading predictions can shape campaign strategies, influence voter turnout, and affect public perception of candidates. When polls consistently miss the mark, it undermines trust in the electoral process and the institutions that support it.

Moving Forward: A Call for Better Polling Practices

To restore confidence in polling, several steps need to be taken.

Enhanced Methodologies. Pollsters must refine their techniques to better capture the diversity of the electorate. This includes improving sampling methods and addressing biases in data collection.

Transparency. Greater transparency in how polls are conducted and how data is adjusted can help the public understand the limitations and potential errors in polling.

Continuous Improvement. Polling organizations should continuously evaluate and update their methodologies based on past performance and emerging trends.

Let´s Resume It!

In conclusion, while economic forecasting will never be an exact science, the repeated failures in Brazil highlight the need for a more thoughtful and accurate approach. Only then can businesses, investors, and the public make truly informed decisions.

While polling will never be an exact science, there is a clear need for more accurate and reliable methods. By addressing the current shortcomings, pollsters can provide more trustworthy insights, ultimately contributing to a healthier democratic process.

But one thing seems very clear. The problem with such poor forecasting results is the fatal combination of an increasingly complex world and ideological prejudices on the part of professionals. On the one hand, it is suggested that linear thinking can capture the many currents in a society and economy, and on the other, this is then dangerously underpinned by the ideological herd mentality of the respective analysts or pollsters.

There are only two things that can help against this: firstly, accept that we are operating in a non-linear environment and secondly, keep your brain safe!

And of course - probably the most important thing of all - think in scenarios!

Have you experienced the impact of inaccurate forecasts in your industry? Share your insights in the comments!